Large Neighborhood Search beyond MIP

Large neighborhood search (LNS) heuristics are an important component of modern branch-and-cut algorithms for solving mixed-integer linear programs (MIPs). Most of these LNS heuristics use the LP relaxation as the basis for their search, which is a reasonable choice in case of MIPs. However, for more general problem classes, the LP relaxation alone may not contain enough information about the original problem to find feasible solutions with these heuristics, e.g., if the problem is nonlinear or not all constraints are present in the current relaxation. In this paper, we discuss a generic way to extend LNS heuristics that have been developed for MIP to constraint integer programming (CIP), which is a generalization of MIP in the direction of constraint programming (CP). We present computational results of LNS heuristics for three problem classes: mixed-integer quadratically constrained programs, nonlinear pseudo-Boolean optimization instances, and resource-constrained project scheduling problems. Therefore, we have implemented extended versions of the following LNS heuristics in the constraint integer programming framework SCIP: Local Branching, RINS, RENS, Crossover, and DINS. Our results indicate that a generic generalization of LNS heuristics to CIP considerably improves the success rate of these heuristics.

[1]  Timo Berthold,et al.  On the computational impact of MIQCP solver components , 2011 .

[2]  Andrea Lodi,et al.  CP-Based Local Branching , 2007, CP.

[3]  Timo Berthold,et al.  Nonlinear Pseudo-Boolean Optimization: Relaxation or Propagation? , 2009, SAT.

[4]  Timo Berthold,et al.  Extending a CIP framework to solve MIQCPs , 2012 .

[5]  Rainer Kolisch,et al.  PSPLIB - A project scheduling problem library: OR Software - ORSEP Operations Research Software Exchange Program , 1997 .

[6]  Michel Gendreau,et al.  A Constraint Programming Framework for Local Search Methods , 1999, J. Heuristics.

[7]  Nicolas Beldiceanu,et al.  Extending CHIP in order to solve complex scheduling and placement problems , 1993, JFPL.

[8]  Edward M. B. Smith,et al.  A symbolic reformulation/spatial branch-and-bound algorithm for the global optimisation of nonconvex MINLPs , 1999 .

[9]  Edward Rothberg,et al.  An Evolutionary Algorithm for Polishing Mixed Integer Programming Solutions , 2007, INFORMS J. Comput..

[10]  David Pisinger,et al.  Large Neighborhood Search , 2018, Handbook of Metaheuristics.

[11]  J. K. Lenstra,et al.  Local Search in Combinatorial Optimisation. , 1997 .

[12]  Timo Berthold,et al.  RENS - Relaxation Enforced Neighborhood Search , 2007 .

[13]  Timo Berthold.  Primal Heuristics for Mixed Integer Programs , 2006 .

[14]  Leo Liberti,et al.  A Local Branching Heuristic for MINLPs , 2008, 0812.2188.

[15]  Tobias Achterberg,et al.  SCIP: solving constraint integer programs , 2009, Math. Program. Comput..

[16]  Thorsten Koch,et al.  Comparing MIQCP Solvers to a Specialised Algorithm for Mine Production Scheduling , 2009, HPSC.

[17]  Tobias Achterberg,et al.  Conflict analysis in mixed integer programming , 2007, Discret. Optim..

[18]  Shubhashis Ghosh,et al.  DINS, a MIP Improvement Heuristic , 2007, IPCO.

[19]  Rolf H. Möhring,et al.  A Constraint Integer Programming Approach for Resource-Constrained Project Scheduling , 2010, CPAIOR.

[20]  Laurence A. Wolsey,et al.  Integer and Combinatorial Optimization , 1988 .

[21]  Avishai Wool,et al.  Algebraic Side-Channel Analysis in the Presence of Errors , 2010, CHES.

[22]  Claude Le Pape,et al.  Exploring relaxation induced neighborhoods to improve MIP solutions , 2005, Math. Program..

[23]  Rolf H. Möhring,et al.  Solving Project Scheduling Problems by Minimum Cut Computations , 2002, Manag. Sci..